Are U.S. coronavirus deaths being over counted or under counted?

It’s frustrating how little we know about the Wuhan coronavirus months after its spread began. We don’t even know with great confidence the approximate number of people who have died from it.

It’s possible to die from the virus without having been tested. It’s also possible to test positive and die due to something else.

John has presented evidence that total deaths in the U.S. are down year-to-date from last year, notwithstanding the coronavirus pandemic. According to this source, they are also down year-to-date from 2018 and 2017.

However, John McCormack notes that, according to the New York Times, there are typically about 6,000 deaths in New York City between March 11 and April 25 in any year, but during that same time period in 2020 there were about 27,000 deaths in the city. The Times also says that the total number of confirmed and probable deaths from the virus is significantly below the 21,000 “excess deaths” the city experienced from mid-March to late-April.

Similarly, the Washington Post tells us that total U.S. deaths surged early in the pandemic. It cites a study by Yale University that found the U.S. recorded an estimated 15,400 excess deaths in “the early weeks of the pandemic” (“the weeks leading up to April 4”) nearly twice as many as were publicly attributed to the Wuhan coronavirus at the time. It’s plausible that deaths from the virus were under counted during this period, given the relatively small number of tests administered as of that time.

The Post notes that the excess deaths aren’t necessarily attributable to the virus itself. They may include people who were afraid to seek medical treatment for unrelated illnesses. And, of course, there can always be fluctuations in the death rate.

An honest study of deaths during periods after the corornavirus took hold is more informative for present purposes than an honest study of deaths year-to-date. However, we’re still in the month of April, and I’m having trouble reconciling a study that shows a large surplus of deaths for multiple weeks with a study that shows a meaningful deficit for the full year.

It’s impossible for me to believe that this virus hasn’t caused a substantial number of deaths that wouldn’t have occurred in its absence. Whether it has caused a significantly larger (or smaller) number of deaths than have been attributed to it is far less clear.

Do officials have incentives to over count coronavirus deaths? I guess so. Some of the counters may be influenced by a desire to see the lockdown continue for many more months. Some officials may want to show that their decision to order a lockdown was justified by the number of deaths. In addition, the government apparently pays more to hospitals if a death is reported as caused by the Wuhan coronavirus.

Do officials have incentives to under count deaths? I guess so. The lower the count in April and thereafter, the more plausibly they can claim their lockdown worked. And, at the end of the day, no sensible governor will want an inflated death count in his or her state. The fewer deaths, the better the governor and his team will look.

How, then, do we stand? Flat-footed, as Groucho Marx would answer.

Responses